多分辨率分析在测井处理及层序划分中的应用  

Application of multi-resolution analysis in logging processing and sequence classification

作　　者：王志坤[1] 钟建华[1] 禚元杰[1] 张磊[1] 

机构地区：[1]中国石油大学资源与信息学院,山东东营257061

出　　处：《大庆石油地质与开发》2008年第3期128-132,共5页Petroleum Geology & Oilfield Development in Daqing

基　　金：中国石油天然气集团公司创新基金(2002F70108)资助

摘　　要：测井是石油地质勘探的有力工具,但测井信号也存在一些缺陷:被噪声污染,信噪比低;纵向分辨率低,无法满足精细的小层对比划分;隐含周期和突变点不够明显,人工判读带有主观随意性。通过小波变换的多分辨率分析可以解决以上难题:小波包消噪处理可以去掉噪声,提高信号的信噪比;离散小波变换对信号多尺度分解,凸显高频部分,重构信号而达到提高测井序列纵向分辨率的目的;对测井信号进行连续小波变换,识别隐含周期和突变点,确定层序界面和湖泛面,为层序划分和地层对比提供定量参考。Well logging is the useful tool of the petroleum geological exploration. But there are several limitations in the logging signal of our application： the SNR of the signal is very low and is contaminated by the noise; with low vertical resolution, it can＇t meet the refined thin layer division and contrast; cryptic period and the mutation point weren＇t so clear, and the manual judgment and explanation is with too much subjective randomness. This difficult problem can be solved by the muhi-resolution analysis of wavelet transformation ： noise discrimination for wavelet package can eliminate the noise, and enhance the SNR; discrete wavelet transformation can multi-scale decompose the signal, on highlighting the high frequency part, and reconstruct signal to achieve the goal of improving the vertical resolution of well logs ; using the continuous wavelet transform to recognize the cryptic period and mutation point and to determine the sequence boundaries and lake flooding surface, and sever the quantitative reference for the sequence division and stratigraphic correlation.

关 键 词：多分辨率分析 测井信号 消噪 层序划分 

分 类 号：P631.811[天文地球—地质矿产勘探]